Voice authentication systems authenticate users based on characteristics of users' voices. Voice authentication systems are designed to recognize and identify the person that is speaking, based on the acoustic features of speech that tend to differ from person to person. In most cases, voice authentication systems authenticate users by comparing a spoken passphrase with a voiceprint of a previously-recorded version of the same passphrase that was generated when voice authentication credentials were first established or later updated.